Dan Cohen, “Virtue and its Discontents”
Virtue-argumentation theory is out to try to answer three evaluative questions about argument by way of focusing on one. The three questions are:
What makes an arguer praiseworthy?
What makes an argument praiseworthy?
What makes argumentation praiseworthy?
The virtue theory takes the answer to the first question to be the means to answer the other two. There are two methodological programs for answering any of these questions:
Top down: argue for an ideal, then apply to cases.
Bottom up: gather empirical data about specific cases, gain theoretical insights.
Cohen’s method is a “weird” or “mixed” methodology. Theorize from “strange cases”: arguing with god, filibusters, academic arguments, impossible arguments, choosing to argue when you shouldn’t, missing arguments, misplaced arguments.  The strategy is to say: here are cases where there are arguments that don’t yield good things, so what is it for arguments to yield something positive? What is a satisfying argument? Roughly, the thick concept of a satisfying argument should be in the right place, the right time, to the right people, and on the right topics (and for the right reasons). That’s a matter of context. Notice that validity is neither considered either a necessary nor sufficient condition for being a good argument. Not sufficient: e.g., “a and b are both P, so a is P” isvalid but not satisfying, and not sufficient: e.g., some arguments are interesting and worthwhile, even if wrongly formed.
So, Cohen poses a sorite: for good arguments, the arguers must argue well. To argue well, they must be good arguers. And to be good arguers, they must have stable habits of mind — good arguments do not happen by accident.